Population-based febrile respiratory illness surveillance was used to monitor the effectiveness of the mandatory adenovirus vaccine administered to all US military trainees. Despite the general success of the vaccination program, surveillance still detected breakthrough infections cases of acute respiratory disease associated with the adenovirus serotypes (4 and 7) specifically targeted by the vaccine. Termination of the vaccine program in 1999 allowed collection of matching samples from an unvaccinated population in the same environment. To explore the possible role of adenoviral coinfection (simultaneous infection by multiple pathogenic adenovirus species) in the facilitation of vaccine breakthrough we compared vaccinated and unvaccinated populations using three independent methods capable of simultaneously detecting multiple adenoviral species: a 70-mer DNA microarray, a commercially available PCR-ELISA, and a multiplex PCR assay. Analysis of 52 patient throat swab samples (21 vaccinated, 31 unvaccinated) collected from 1996-2000 revealed that all vaccinated samples harbored adenoviral coinfections. Most of the coinfections were composed of community-acquired, pathogenic serotypes of both species B (serotypes 3, 7, and/or 21) and species E (serotype 4). Unvaccinated samples primarily contained just serotype 4. Coinfections may be selected in vaccinated individuals because they can produce mosaics with unique antigenicity, giving them the ability to bypass vaccine-induced immunity specific to virus particles grown in isolation. This study highlights the previously undocumented phenomenon of natural adenoviral coinfections in an environment suitable for the generation of new, potentially virulent, and uniquely selectable recombinational variants, and suggests that vaccination may play a role in coinfection.